u-boot/tools/patman/patchstream.py

646 lines
23 KiB
Python
Raw Normal View History

# SPDX-License-Identifier: GPL-2.0+
# Copyright (c) 2011 The Chromium OS Authors.
#
patman: Use the Change-Id, version, and prefix in the Message-Id As per the centithread on ksummit-discuss [1], there are folks who feel that if a Change-Id is present in a developer's local commit that said Change-Id could be interesting to include in upstream posts. Specifically if two commits are posted with the same Change-Id there's a reasonable chance that they are either the same commit or a newer version of the same commit. Specifically this is because that's how gerrit has trained people to work. There is much angst about Change-Id in upstream Linux, but one thing that seems safe and non-controversial is to include the Change-Id as part of the string of crud that makes up a Message-Id. Let's give that a try. In theory (if there is enough adoption) this could help a tool more reliably find various versions of a commit. This actually might work pretty well for U-Boot where (I believe) quite a number of developers use patman, so there could be critical mass (assuming that enough of these people also use a git hook that adds Change-Id to their commits). I was able to find this git hook by searching for "gerrit change id git hook" in my favorite search engine. In theory one could imagine something like this could be integrated into other tools, possibly even git-send-email. Getting it into patman seems like a sane first step, though. NOTE: this patch is being posted using a patman containing this patch, so you should be able to see the Message-Id of this patch and see that it contains my local Change-Id, which ends in 2b9 if you want to check. [1] https://lists.linuxfoundation.org/pipermail/ksummit-discuss/2019-August/006739.html Signed-off-by: Douglas Anderson <dianders@chromium.org>
2019-09-27 16:23:56 +00:00
import datetime
import math
import os
import re
import shutil
import tempfile
from patman import command
from patman import commit
from patman import gitutil
from patman.series import Series
# Tags that we detect and remove
patman: Use the Change-Id, version, and prefix in the Message-Id As per the centithread on ksummit-discuss [1], there are folks who feel that if a Change-Id is present in a developer's local commit that said Change-Id could be interesting to include in upstream posts. Specifically if two commits are posted with the same Change-Id there's a reasonable chance that they are either the same commit or a newer version of the same commit. Specifically this is because that's how gerrit has trained people to work. There is much angst about Change-Id in upstream Linux, but one thing that seems safe and non-controversial is to include the Change-Id as part of the string of crud that makes up a Message-Id. Let's give that a try. In theory (if there is enough adoption) this could help a tool more reliably find various versions of a commit. This actually might work pretty well for U-Boot where (I believe) quite a number of developers use patman, so there could be critical mass (assuming that enough of these people also use a git hook that adds Change-Id to their commits). I was able to find this git hook by searching for "gerrit change id git hook" in my favorite search engine. In theory one could imagine something like this could be integrated into other tools, possibly even git-send-email. Getting it into patman seems like a sane first step, though. NOTE: this patch is being posted using a patman containing this patch, so you should be able to see the Message-Id of this patch and see that it contains my local Change-Id, which ends in 2b9 if you want to check. [1] https://lists.linuxfoundation.org/pipermail/ksummit-discuss/2019-August/006739.html Signed-off-by: Douglas Anderson <dianders@chromium.org>
2019-09-27 16:23:56 +00:00
re_remove = re.compile('^BUG=|^TEST=|^BRANCH=|^Review URL:'
'|Reviewed-on:|Commit-\w*:')
# Lines which are allowed after a TEST= line
re_allowed_after_test = re.compile('^Signed-off-by:')
# Signoffs
re_signoff = re.compile('^Signed-off-by: *(.*)')
2020-05-04 20:28:34 +00:00
# Cover letter tag
re_cover = re.compile('^Cover-([a-z-]*): *(.*)')
# Patch series tag
re_series_tag = re.compile('^Series-([a-z-]*): *(.*)')
patman: Use the Change-Id, version, and prefix in the Message-Id As per the centithread on ksummit-discuss [1], there are folks who feel that if a Change-Id is present in a developer's local commit that said Change-Id could be interesting to include in upstream posts. Specifically if two commits are posted with the same Change-Id there's a reasonable chance that they are either the same commit or a newer version of the same commit. Specifically this is because that's how gerrit has trained people to work. There is much angst about Change-Id in upstream Linux, but one thing that seems safe and non-controversial is to include the Change-Id as part of the string of crud that makes up a Message-Id. Let's give that a try. In theory (if there is enough adoption) this could help a tool more reliably find various versions of a commit. This actually might work pretty well for U-Boot where (I believe) quite a number of developers use patman, so there could be critical mass (assuming that enough of these people also use a git hook that adds Change-Id to their commits). I was able to find this git hook by searching for "gerrit change id git hook" in my favorite search engine. In theory one could imagine something like this could be integrated into other tools, possibly even git-send-email. Getting it into patman seems like a sane first step, though. NOTE: this patch is being posted using a patman containing this patch, so you should be able to see the Message-Id of this patch and see that it contains my local Change-Id, which ends in 2b9 if you want to check. [1] https://lists.linuxfoundation.org/pipermail/ksummit-discuss/2019-August/006739.html Signed-off-by: Douglas Anderson <dianders@chromium.org>
2019-09-27 16:23:56 +00:00
# Change-Id will be used to generate the Message-Id and then be stripped
re_change_id = re.compile('^Change-Id: *(.*)')
# Commit series tag
re_commit_tag = re.compile('^Commit-([a-z-]*): *(.*)')
# Commit tags that we want to collect and keep
re_tag = re.compile('^(Tested-by|Acked-by|Reviewed-by|Patch-cc|Fixes): (.*)')
# The start of a new commit in the git log
re_commit = re.compile('^commit ([0-9a-f]*)$')
# We detect these since checkpatch doesn't always do it
re_space_before_tab = re.compile('^[+].* \t')
# Match indented lines for changes
re_leading_whitespace = re.compile('^\s')
# States we can be in - can we use range() and still have comments?
STATE_MSG_HEADER = 0 # Still in the message header
STATE_PATCH_SUBJECT = 1 # In patch subject (first line of log for a commit)
STATE_PATCH_HEADER = 2 # In patch header (after the subject)
STATE_DIFFS = 3 # In the diff part (past --- line)
class PatchStream:
"""Class for detecting/injecting tags in a patch or series of patches
We support processing the output of 'git log' to read out the tags we
are interested in. We can also process a patch file in order to remove
unwanted tags or inject additional ones. These correspond to the two
phases of processing.
"""
def __init__(self, series, name=None, is_log=False):
self.skip_blank = False # True to skip a single blank line
self.found_test = False # Found a TEST= line
2020-05-04 20:28:34 +00:00
self.lines_after_test = 0 # Number of lines found after TEST=
self.warn = [] # List of warnings we have collected
self.linenum = 1 # Output line number we are up to
self.in_section = None # Name of start...END section we are in
self.notes = [] # Series notes
self.section = [] # The current section...END section
self.series = series # Info about the patch series
self.is_log = is_log # True if indent like git log
2020-05-04 20:28:34 +00:00
self.in_change = None # Name of the change list we are in
self.change_version = 0 # Non-zero if we are in a change list
self.change_lines = [] # Lines of the current change
self.blank_count = 0 # Number of blank lines stored up
self.state = STATE_MSG_HEADER # What state are we in?
self.signoff = [] # Contents of signoff line
self.commit = None # Current commit
def AddToSeries(self, line, name, value):
"""Add a new Series-xxx tag.
When a Series-xxx tag is detected, we come here to record it, if we
are scanning a 'git log'.
Args:
line: Source line containing tag (useful for debug/error messages)
name: Tag name (part after 'Series-')
value: Tag value (part after 'Series-xxx: ')
"""
if name == 'notes':
self.in_section = name
self.skip_blank = False
if self.is_log:
self.series.AddTag(self.commit, line, name, value)
def AddToCommit(self, line, name, value):
"""Add a new Commit-xxx tag.
When a Commit-xxx tag is detected, we come here to record it.
Args:
line: Source line containing tag (useful for debug/error messages)
name: Tag name (part after 'Commit-')
value: Tag value (part after 'Commit-xxx: ')
"""
if name == 'notes':
self.in_section = 'commit-' + name
self.skip_blank = False
def AddCommitRtag(self, rtag_type, who):
"""Add a response tag to the current commit
Args:
key: rtag type (e.g. 'Reviewed-by')
who: Person who gave that rtag, e.g. 'Fred Bloggs <fred@bloggs.org>'
"""
self.commit.AddRtag(rtag_type, who)
def CloseCommit(self):
"""Save the current commit into our commit list, and reset our state"""
if self.commit and self.is_log:
self.series.AddCommit(self.commit)
self.commit = None
# If 'END' is missing in a 'Cover-letter' section, and that section
# happens to show up at the very end of the commit message, this is
# the chance for us to fix it up.
if self.in_section == 'cover' and self.is_log:
self.series.cover = self.section
self.in_section = None
self.skip_blank = True
self.section = []
2020-05-04 20:28:34 +00:00
def ParseVersion(self, value, line):
"""Parse a version from a *-changes tag
Args:
value: Tag value (part after 'xxx-changes: '
line: Source line containing tag
Returns:
The version as an integer
"""
try:
return int(value)
except ValueError as str:
raise ValueError("%s: Cannot decode version info '%s'" %
(self.commit.hash, line))
def FinalizeChange(self):
"""Finalize a (multi-line) change and add it to the series or commit"""
if not self.change_lines:
return
change = '\n'.join(self.change_lines)
if self.in_change == 'Series':
self.series.AddChange(self.change_version, self.commit, change)
elif self.in_change == 'Cover':
self.series.AddChange(self.change_version, None, change)
elif self.in_change == 'Commit':
self.commit.AddChange(self.change_version, change)
self.change_lines = []
def ProcessLine(self, line):
"""Process a single line of a patch file or commit log
This process a line and returns a list of lines to output. The list
may be empty or may contain multiple output lines.
This is where all the complicated logic is located. The class's
state is used to move between different states and detect things
properly.
We can be in one of two modes:
self.is_log == True: This is 'git log' mode, where most output is
indented by 4 characters and we are scanning for tags
self.is_log == False: This is 'patch' mode, where we already have
all the tags, and are processing patches to remove junk we
don't want, and add things we think are required.
Args:
line: text line to process
Returns:
list of output lines, or [] if nothing should be output
"""
# Initially we have no output. Prepare the input line string
out = []
line = line.rstrip('\n')
commit_match = re_commit.match(line) if self.is_log else None
if self.is_log:
if line[:4] == ' ':
line = line[4:]
# Handle state transition and skipping blank lines
series_tag_match = re_series_tag.match(line)
patman: Use the Change-Id, version, and prefix in the Message-Id As per the centithread on ksummit-discuss [1], there are folks who feel that if a Change-Id is present in a developer's local commit that said Change-Id could be interesting to include in upstream posts. Specifically if two commits are posted with the same Change-Id there's a reasonable chance that they are either the same commit or a newer version of the same commit. Specifically this is because that's how gerrit has trained people to work. There is much angst about Change-Id in upstream Linux, but one thing that seems safe and non-controversial is to include the Change-Id as part of the string of crud that makes up a Message-Id. Let's give that a try. In theory (if there is enough adoption) this could help a tool more reliably find various versions of a commit. This actually might work pretty well for U-Boot where (I believe) quite a number of developers use patman, so there could be critical mass (assuming that enough of these people also use a git hook that adds Change-Id to their commits). I was able to find this git hook by searching for "gerrit change id git hook" in my favorite search engine. In theory one could imagine something like this could be integrated into other tools, possibly even git-send-email. Getting it into patman seems like a sane first step, though. NOTE: this patch is being posted using a patman containing this patch, so you should be able to see the Message-Id of this patch and see that it contains my local Change-Id, which ends in 2b9 if you want to check. [1] https://lists.linuxfoundation.org/pipermail/ksummit-discuss/2019-August/006739.html Signed-off-by: Douglas Anderson <dianders@chromium.org>
2019-09-27 16:23:56 +00:00
change_id_match = re_change_id.match(line)
commit_tag_match = re_commit_tag.match(line)
cover_match = re_cover.match(line)
signoff_match = re_signoff.match(line)
leading_whitespace_match = re_leading_whitespace.match(line)
tag_match = None
if self.state == STATE_PATCH_HEADER:
tag_match = re_tag.match(line)
is_blank = not line.strip()
if is_blank:
if (self.state == STATE_MSG_HEADER
or self.state == STATE_PATCH_SUBJECT):
self.state += 1
# We don't have a subject in the text stream of patch files
# It has its own line with a Subject: tag
if not self.is_log and self.state == STATE_PATCH_SUBJECT:
self.state += 1
elif commit_match:
self.state = STATE_MSG_HEADER
# If a tag is detected, or a new commit starts
patman: Use the Change-Id, version, and prefix in the Message-Id As per the centithread on ksummit-discuss [1], there are folks who feel that if a Change-Id is present in a developer's local commit that said Change-Id could be interesting to include in upstream posts. Specifically if two commits are posted with the same Change-Id there's a reasonable chance that they are either the same commit or a newer version of the same commit. Specifically this is because that's how gerrit has trained people to work. There is much angst about Change-Id in upstream Linux, but one thing that seems safe and non-controversial is to include the Change-Id as part of the string of crud that makes up a Message-Id. Let's give that a try. In theory (if there is enough adoption) this could help a tool more reliably find various versions of a commit. This actually might work pretty well for U-Boot where (I believe) quite a number of developers use patman, so there could be critical mass (assuming that enough of these people also use a git hook that adds Change-Id to their commits). I was able to find this git hook by searching for "gerrit change id git hook" in my favorite search engine. In theory one could imagine something like this could be integrated into other tools, possibly even git-send-email. Getting it into patman seems like a sane first step, though. NOTE: this patch is being posted using a patman containing this patch, so you should be able to see the Message-Id of this patch and see that it contains my local Change-Id, which ends in 2b9 if you want to check. [1] https://lists.linuxfoundation.org/pipermail/ksummit-discuss/2019-August/006739.html Signed-off-by: Douglas Anderson <dianders@chromium.org>
2019-09-27 16:23:56 +00:00
if series_tag_match or commit_tag_match or change_id_match or \
2020-05-04 20:28:34 +00:00
cover_match or signoff_match or self.state == STATE_MSG_HEADER:
# but we are already in a section, this means 'END' is missing
# for that section, fix it up.
if self.in_section:
self.warn.append("Missing 'END' in section '%s'" % self.in_section)
if self.in_section == 'cover':
self.series.cover = self.section
elif self.in_section == 'notes':
if self.is_log:
self.series.notes += self.section
elif self.in_section == 'commit-notes':
if self.is_log:
self.commit.notes += self.section
else:
self.warn.append("Unknown section '%s'" % self.in_section)
self.in_section = None
self.skip_blank = True
self.section = []
# but we are already in a change list, that means a blank line
# is missing, fix it up.
if self.in_change:
2020-05-04 20:28:34 +00:00
self.warn.append("Missing 'blank line' in section '%s-changes'" % self.in_change)
self.FinalizeChange()
2020-05-04 20:28:34 +00:00
self.in_change = None
self.change_version = 0
# If we are in a section, keep collecting lines until we see END
if self.in_section:
if line == 'END':
if self.in_section == 'cover':
self.series.cover = self.section
elif self.in_section == 'notes':
if self.is_log:
self.series.notes += self.section
elif self.in_section == 'commit-notes':
if self.is_log:
self.commit.notes += self.section
else:
self.warn.append("Unknown section '%s'" % self.in_section)
self.in_section = None
self.skip_blank = True
self.section = []
else:
self.section.append(line)
# If we are not in a section, it is an unexpected END
elif line == 'END':
raise ValueError("'END' wihout section")
# Detect the commit subject
elif not is_blank and self.state == STATE_PATCH_SUBJECT:
self.commit.subject = line
# Detect the tags we want to remove, and skip blank lines
elif re_remove.match(line) and not commit_tag_match:
self.skip_blank = True
# TEST= should be the last thing in the commit, so remove
# everything after it
if line.startswith('TEST='):
self.found_test = True
elif self.skip_blank and is_blank:
self.skip_blank = False
2020-05-04 20:28:34 +00:00
# Detect Cover-xxx tags
elif cover_match:
2020-05-04 20:28:34 +00:00
name = cover_match.group(1)
value = cover_match.group(2)
if name == 'letter':
self.in_section = 'cover'
self.skip_blank = False
elif name == 'letter-cc':
self.AddToSeries(line, 'cover-cc', value)
elif name == 'changes':
self.in_change = 'Cover'
self.change_version = self.ParseVersion(value, line)
# If we are in a change list, key collected lines until a blank one
elif self.in_change:
if is_blank:
# Blank line ends this change list
self.FinalizeChange()
2020-05-04 20:28:34 +00:00
self.in_change = None
self.change_version = 0
elif line == '---':
self.FinalizeChange()
2020-05-04 20:28:34 +00:00
self.in_change = None
self.change_version = 0
out = self.ProcessLine(line)
elif self.is_log:
if not leading_whitespace_match:
self.FinalizeChange()
self.change_lines.append(line)
self.skip_blank = False
# Detect Series-xxx tags
elif series_tag_match:
name = series_tag_match.group(1)
value = series_tag_match.group(2)
if name == 'changes':
# value is the version number: e.g. 1, or 2
2020-05-04 20:28:34 +00:00
self.in_change = 'Series'
self.change_version = self.ParseVersion(value, line)
else:
self.AddToSeries(line, name, value)
self.skip_blank = True
patman: Use the Change-Id, version, and prefix in the Message-Id As per the centithread on ksummit-discuss [1], there are folks who feel that if a Change-Id is present in a developer's local commit that said Change-Id could be interesting to include in upstream posts. Specifically if two commits are posted with the same Change-Id there's a reasonable chance that they are either the same commit or a newer version of the same commit. Specifically this is because that's how gerrit has trained people to work. There is much angst about Change-Id in upstream Linux, but one thing that seems safe and non-controversial is to include the Change-Id as part of the string of crud that makes up a Message-Id. Let's give that a try. In theory (if there is enough adoption) this could help a tool more reliably find various versions of a commit. This actually might work pretty well for U-Boot where (I believe) quite a number of developers use patman, so there could be critical mass (assuming that enough of these people also use a git hook that adds Change-Id to their commits). I was able to find this git hook by searching for "gerrit change id git hook" in my favorite search engine. In theory one could imagine something like this could be integrated into other tools, possibly even git-send-email. Getting it into patman seems like a sane first step, though. NOTE: this patch is being posted using a patman containing this patch, so you should be able to see the Message-Id of this patch and see that it contains my local Change-Id, which ends in 2b9 if you want to check. [1] https://lists.linuxfoundation.org/pipermail/ksummit-discuss/2019-August/006739.html Signed-off-by: Douglas Anderson <dianders@chromium.org>
2019-09-27 16:23:56 +00:00
# Detect Change-Id tags
elif change_id_match:
value = change_id_match.group(1)
if self.is_log:
if self.commit.change_id:
raise ValueError("%s: Two Change-Ids: '%s' vs. '%s'" %
(self.commit.hash, self.commit.change_id, value))
self.commit.change_id = value
self.skip_blank = True
# Detect Commit-xxx tags
elif commit_tag_match:
name = commit_tag_match.group(1)
value = commit_tag_match.group(2)
if name == 'notes':
self.AddToCommit(line, name, value)
self.skip_blank = True
2020-05-04 20:28:34 +00:00
elif name == 'changes':
self.in_change = 'Commit'
self.change_version = self.ParseVersion(value, line)
# Detect the start of a new commit
elif commit_match:
self.CloseCommit()
self.commit = commit.Commit(commit_match.group(1))
# Detect tags in the commit message
elif tag_match:
rtag_type, who = tag_match.groups()
self.AddCommitRtag(rtag_type, who)
# Remove Tested-by self, since few will take much notice
if (rtag_type == 'Tested-by' and
who.find(os.getenv('USER') + '@') != -1):
self.warn.append("Ignoring %s" % line)
elif rtag_type == 'Patch-cc':
self.commit.AddCc(who.split(','))
else:
out = [line]
# Suppress duplicate signoffs
elif signoff_match:
if (self.is_log or not self.commit or
self.commit.CheckDuplicateSignoff(signoff_match.group(1))):
out = [line]
# Well that means this is an ordinary line
else:
# Look for space before tab
m = re_space_before_tab.match(line)
if m:
self.warn.append('Line %d/%d has space before tab' %
(self.linenum, m.start()))
# OK, we have a valid non-blank line
out = [line]
self.linenum += 1
self.skip_blank = False
if self.state == STATE_DIFFS:
pass
# If this is the start of the diffs section, emit our tags and
# change log
elif line == '---':
self.state = STATE_DIFFS
2020-05-04 20:28:34 +00:00
# Output the tags (signoff first), then change list
out = []
log = self.series.MakeChangeLog(self.commit)
out += [line]
if self.commit:
out += self.commit.notes
out += [''] + log
elif self.found_test:
if not re_allowed_after_test.match(line):
self.lines_after_test += 1
return out
def Finalize(self):
"""Close out processing of this patch stream"""
self.FinalizeChange()
self.CloseCommit()
if self.lines_after_test:
self.warn.append('Found %d lines after TEST=' %
self.lines_after_test)
patman: Use the Change-Id, version, and prefix in the Message-Id As per the centithread on ksummit-discuss [1], there are folks who feel that if a Change-Id is present in a developer's local commit that said Change-Id could be interesting to include in upstream posts. Specifically if two commits are posted with the same Change-Id there's a reasonable chance that they are either the same commit or a newer version of the same commit. Specifically this is because that's how gerrit has trained people to work. There is much angst about Change-Id in upstream Linux, but one thing that seems safe and non-controversial is to include the Change-Id as part of the string of crud that makes up a Message-Id. Let's give that a try. In theory (if there is enough adoption) this could help a tool more reliably find various versions of a commit. This actually might work pretty well for U-Boot where (I believe) quite a number of developers use patman, so there could be critical mass (assuming that enough of these people also use a git hook that adds Change-Id to their commits). I was able to find this git hook by searching for "gerrit change id git hook" in my favorite search engine. In theory one could imagine something like this could be integrated into other tools, possibly even git-send-email. Getting it into patman seems like a sane first step, though. NOTE: this patch is being posted using a patman containing this patch, so you should be able to see the Message-Id of this patch and see that it contains my local Change-Id, which ends in 2b9 if you want to check. [1] https://lists.linuxfoundation.org/pipermail/ksummit-discuss/2019-August/006739.html Signed-off-by: Douglas Anderson <dianders@chromium.org>
2019-09-27 16:23:56 +00:00
def WriteMessageId(self, outfd):
"""Write the Message-Id into the output.
This is based on the Change-Id in the original patch, the version,
and the prefix.
Args:
outfd: Output stream file object
"""
if not self.commit.change_id:
return
# If the count is -1 we're testing, so use a fixed time
if self.commit.count == -1:
time_now = datetime.datetime(1999, 12, 31, 23, 59, 59)
else:
time_now = datetime.datetime.now()
# In theory there is email.utils.make_msgid() which would be nice
# to use, but it already produces something way too long and thus
# will produce ugly commit lines if someone throws this into
# a "Link:" tag in the final commit. So (sigh) roll our own.
# Start with the time; presumably we wouldn't send the same series
# with the same Change-Id at the exact same second.
parts = [time_now.strftime("%Y%m%d%H%M%S")]
# These seem like they would be nice to include.
if 'prefix' in self.series:
parts.append(self.series['prefix'])
if 'version' in self.series:
parts.append("v%s" % self.series['version'])
parts.append(str(self.commit.count + 1))
# The Change-Id must be last, right before the @
parts.append(self.commit.change_id)
# Join parts together with "." and write it out.
outfd.write('Message-Id: <%s@changeid>\n' % '.'.join(parts))
def ProcessStream(self, infd, outfd):
"""Copy a stream from infd to outfd, filtering out unwanting things.
This is used to process patch files one at a time.
Args:
infd: Input stream file object
outfd: Output stream file object
"""
# Extract the filename from each diff, for nice warnings
fname = None
last_fname = None
re_fname = re.compile('diff --git a/(.*) b/.*')
patman: Use the Change-Id, version, and prefix in the Message-Id As per the centithread on ksummit-discuss [1], there are folks who feel that if a Change-Id is present in a developer's local commit that said Change-Id could be interesting to include in upstream posts. Specifically if two commits are posted with the same Change-Id there's a reasonable chance that they are either the same commit or a newer version of the same commit. Specifically this is because that's how gerrit has trained people to work. There is much angst about Change-Id in upstream Linux, but one thing that seems safe and non-controversial is to include the Change-Id as part of the string of crud that makes up a Message-Id. Let's give that a try. In theory (if there is enough adoption) this could help a tool more reliably find various versions of a commit. This actually might work pretty well for U-Boot where (I believe) quite a number of developers use patman, so there could be critical mass (assuming that enough of these people also use a git hook that adds Change-Id to their commits). I was able to find this git hook by searching for "gerrit change id git hook" in my favorite search engine. In theory one could imagine something like this could be integrated into other tools, possibly even git-send-email. Getting it into patman seems like a sane first step, though. NOTE: this patch is being posted using a patman containing this patch, so you should be able to see the Message-Id of this patch and see that it contains my local Change-Id, which ends in 2b9 if you want to check. [1] https://lists.linuxfoundation.org/pipermail/ksummit-discuss/2019-August/006739.html Signed-off-by: Douglas Anderson <dianders@chromium.org>
2019-09-27 16:23:56 +00:00
self.WriteMessageId(outfd)
while True:
line = infd.readline()
if not line:
break
out = self.ProcessLine(line)
# Try to detect blank lines at EOF
for line in out:
match = re_fname.match(line)
if match:
last_fname = fname
fname = match.group(1)
if line == '+':
self.blank_count += 1
else:
if self.blank_count and (line == '-- ' or match):
self.warn.append("Found possible blank line(s) at "
"end of file '%s'" % last_fname)
outfd.write('+\n' * self.blank_count)
outfd.write(line + '\n')
self.blank_count = 0
self.Finalize()
def GetMetaDataForList(commit_range, git_dir=None, count=None,
series = None, allow_overwrite=False):
"""Reads out patch series metadata from the commits
This does a 'git log' on the relevant commits and pulls out the tags we
are interested in.
Args:
commit_range: Range of commits to count (e.g. 'HEAD..base')
git_dir: Path to git repositiory (None to use default)
count: Number of commits to list, or None for no limit
series: Series object to add information into. By default a new series
is started.
allow_overwrite: Allow tags to overwrite an existing tag
Returns:
A Series object containing information about the commits.
"""
if not series:
series = Series()
series.allow_overwrite = allow_overwrite
params = gitutil.LogCmd(commit_range, reverse=True, count=count,
git_dir=git_dir)
stdout = command.RunPipe([params], capture=True).stdout
ps = PatchStream(series, is_log=True)
for line in stdout.splitlines():
ps.ProcessLine(line)
ps.Finalize()
return series
def GetMetaData(branch, start, count):
"""Reads out patch series metadata from the commits
This does a 'git log' on the relevant commits and pulls out the tags we
are interested in.
Args:
branch: Branch to use (None for current branch)
start: Commit to start from: 0=branch HEAD, 1=next one, etc.
count: Number of commits to list
"""
return GetMetaDataForList('%s~%d' % (branch if branch else 'HEAD', start),
None, count)
def GetMetaDataForTest(text):
"""Process metadata from a file containing a git log. Used for tests
Args:
text:
"""
series = Series()
ps = PatchStream(series, is_log=True)
for line in text.splitlines():
ps.ProcessLine(line)
ps.Finalize()
return series
def FixPatch(backup_dir, fname, series, commit):
"""Fix up a patch file, by adding/removing as required.
We remove our tags from the patch file, insert changes lists, etc.
The patch file is processed in place, and overwritten.
A backup file is put into backup_dir (if not None).
Args:
fname: Filename to patch file to process
series: Series information about this patch set
commit: Commit object for this patch file
Return:
A list of errors, or [] if all ok.
"""
handle, tmpname = tempfile.mkstemp()
outfd = os.fdopen(handle, 'w', encoding='utf-8')
infd = open(fname, 'r', encoding='utf-8')
ps = PatchStream(series)
ps.commit = commit
ps.ProcessStream(infd, outfd)
infd.close()
outfd.close()
# Create a backup file if required
if backup_dir:
shutil.copy(fname, os.path.join(backup_dir, os.path.basename(fname)))
shutil.move(tmpname, fname)
return ps.warn
def FixPatches(series, fnames):
"""Fix up a list of patches identified by filenames
The patch files are processed in place, and overwritten.
Args:
series: The series object
fnames: List of patch files to process
"""
# Current workflow creates patches, so we shouldn't need a backup
backup_dir = None #tempfile.mkdtemp('clean-patch')
count = 0
for fname in fnames:
commit = series.commits[count]
commit.patch = fname
patman: Use the Change-Id, version, and prefix in the Message-Id As per the centithread on ksummit-discuss [1], there are folks who feel that if a Change-Id is present in a developer's local commit that said Change-Id could be interesting to include in upstream posts. Specifically if two commits are posted with the same Change-Id there's a reasonable chance that they are either the same commit or a newer version of the same commit. Specifically this is because that's how gerrit has trained people to work. There is much angst about Change-Id in upstream Linux, but one thing that seems safe and non-controversial is to include the Change-Id as part of the string of crud that makes up a Message-Id. Let's give that a try. In theory (if there is enough adoption) this could help a tool more reliably find various versions of a commit. This actually might work pretty well for U-Boot where (I believe) quite a number of developers use patman, so there could be critical mass (assuming that enough of these people also use a git hook that adds Change-Id to their commits). I was able to find this git hook by searching for "gerrit change id git hook" in my favorite search engine. In theory one could imagine something like this could be integrated into other tools, possibly even git-send-email. Getting it into patman seems like a sane first step, though. NOTE: this patch is being posted using a patman containing this patch, so you should be able to see the Message-Id of this patch and see that it contains my local Change-Id, which ends in 2b9 if you want to check. [1] https://lists.linuxfoundation.org/pipermail/ksummit-discuss/2019-August/006739.html Signed-off-by: Douglas Anderson <dianders@chromium.org>
2019-09-27 16:23:56 +00:00
commit.count = count
result = FixPatch(backup_dir, fname, series, commit)
if result:
print('%d warnings for %s:' % (len(result), fname))
for warn in result:
print('\t', warn)
print
count += 1
print('Cleaned %d patches' % count)
def InsertCoverLetter(fname, series, count):
"""Inserts a cover letter with the required info into patch 0
Args:
fname: Input / output filename of the cover letter file
series: Series object
count: Number of patches in the series
"""
fd = open(fname, 'r')
lines = fd.readlines()
fd.close()
fd = open(fname, 'w')
text = series.cover
prefix = series.GetPatchPrefix()
for line in lines:
if line.startswith('Subject:'):
# if more than 10 or 100 patches, it should say 00/xx, 000/xxx, etc
zero_repeat = int(math.log10(count)) + 1
zero = '0' * zero_repeat
line = 'Subject: [%s %s/%d] %s\n' % (prefix, zero, count, text[0])
# Insert our cover letter
elif line.startswith('*** BLURB HERE ***'):
# First the blurb test
line = '\n'.join(text[1:]) + '\n'
if series.get('notes'):
line += '\n'.join(series.notes) + '\n'
# Now the change list
out = series.MakeChangeLog(None)
line += '\n' + '\n'.join(out)
fd.write(line)
fd.close()